The mechanical vibration source classifier is a modern library for classifying events around an optical fiber, which is programmed in Python. This library provides the ability to classify events using sophisticated signal processing and machine learning methods. The library also contains a simple API (so-called “application programming interface”). It allows you to easily update the functionality and settings of the parameterization, but also the classification itself based on the configuration file. The package also provides the ability to easily add / remove events for classification, making it flexible and easily extensible.

The desktop application for the MS Windows operating system is used to display the results of activity detection and information about the status of current measurements. It is written in C# and uses the .NET Framework 4.7.2. The application connects to the access MQTT server to which it is subscribed. Detected activities are displayed in a tabular list. It is possible to view and watch activities live on the map. In addition, it is possible to view the history of the activity trajectory and the course of the instantaneous speed over time.

A software system that enables long-term transmission and storage of large amounts of uncompressed audio data to a remote server. The data is transmitted using a TCP stream, while the expected size of the data flow is up to 200MB/s, which in 24/7 operation corresponds to a total size of about 17TB per day. In order to be able to archive such high volumes of data for a long time, three-stage dynamic lossy compression technique is integrated within the system with selectable parameters of individual compression stages.

Features:

  • Independent and reliable system,
  • planned operation in 24/7 mode,
  • continuous incoming data flow in real time,
  • data flow processing independent of other system functionalities,
  • daily registration up to 17 TB, max. annual registration 6205 TB,
  • dynamic backup of the incoming data stream for the last x h.
  • support for multi-threaded processing of large amounts of data,
  • performing dynamic data backup at the level of detected events,
  • creating a long history of events.

 

Supported operating systems: 64-bit version of Windows 7/10, Ubuntu

Komprese